ci.yml

name: CI

on:
  workflow_dispatch:
  push:
    branches:
      - main
  pull_request:
    branches:
      - main

permissions:
  contents: read

concurrency:
  group: ci-${{ github.workflow }}-${{ github.ref }}
  cancel-in-progress: true

jobs:
  python-reference:
    name: Python Reference Checks
    runs-on: ubuntu-latest

    steps:
      - name: Checkout
        uses: actions/checkout@v4

      - name: Setup Python
        uses: actions/setup-python@v5
        with:
          python-version: "3.12"

      - name: Run spec fixtures
        run: make test-spec

      - name: Run Python CLI tests
        run: make test-python

      - name: Validate Python module syntax
        run: python -m compileall diffly-python

      - name: Run CLI smoke test
        run: |
          python diffly-python/diffly.py \
            --a diffly-spec/fixtures/keyed_basic_add_remove_change/a.csv \
            --b diffly-spec/fixtures/keyed_basic_add_remove_change/b.csv \
            --key id > /tmp/diffly-smoke.jsonl

      - name: Run positional CLI smoke test (default mode)
        run: |
          python diffly-python/diffly.py \
            --a diffly-spec/fixtures/positional_basic_add_remove_change/a.csv \
            --b diffly-spec/fixtures/positional_basic_add_remove_change/b.csv > /tmp/diffly-smoke-positional.jsonl

      - name: Run unordered CLI smoke test (--ignore-row-order)
        run: |
          python diffly-python/diffly.py \
            --a diffly-spec/fixtures/positional_ignore_row_order_basic_add_remove/a.csv \
            --b diffly-spec/fixtures/positional_ignore_row_order_basic_add_remove/b.csv \
            --ignore-row-order > /tmp/diffly-smoke-unordered.jsonl

      - name: Run ignore-column-order keyed smoke test
        run: |
          python diffly-python/diffly.py \
            --a diffly-spec/fixtures/keyed_header_sorted_mode_add/a.csv \
            --b diffly-spec/fixtures/keyed_header_sorted_mode_add/b.csv \
            --key id \
            --ignore-column-order > /tmp/diffly-smoke-ignore-column-order.jsonl

      - name: Verify smoke output has schema and stats
        run: |
          python - <<'PY'
          import json
          from pathlib import Path

          lines = [
              json.loads(line)
              for line in Path("/tmp/diffly-smoke.jsonl").read_text(encoding="utf-8").splitlines()
              if line.strip()
          ]
          assert lines, "Smoke output is empty"
          assert lines[0].get("type") == "schema", "First event must be schema"
          assert lines[-1].get("type") == "stats", "Last event must be stats"

          positional_lines = [
              json.loads(line)
              for line in Path("/tmp/diffly-smoke-positional.jsonl").read_text(encoding="utf-8").splitlines()
              if line.strip()
          ]
          assert positional_lines, "Positional smoke output is empty"
          assert positional_lines[0].get("type") == "schema", "First positional event must be schema"
          assert positional_lines[-1].get("type") == "stats", "Last positional event must be stats"

          unordered_lines = [
              json.loads(line)
              for line in Path("/tmp/diffly-smoke-unordered.jsonl").read_text(encoding="utf-8").splitlines()
              if line.strip()
          ]
          assert unordered_lines, "Unordered smoke output is empty"
          assert unordered_lines[0].get("type") == "schema", "First unordered event must be schema"
          assert unordered_lines[-1].get("type") == "stats", "Last unordered event must be stats"
          assert unordered_lines[-1].get("rows_changed") == 0, "Unordered stats rows_changed must be 0"

          ignore_column_lines = [
              json.loads(line)
              for line in Path("/tmp/diffly-smoke-ignore-column-order.jsonl").read_text(encoding="utf-8").splitlines()
              if line.strip()
          ]
          assert ignore_column_lines, "Ignore-column-order smoke output is empty"
          assert ignore_column_lines[0].get("type") == "schema", "First ignore-column-order event must be schema"
          assert ignore_column_lines[-1].get("type") == "stats", "Last ignore-column-order event must be stats"
          PY

  rust-conformance:
    name: Rust Conformance Checks
    runs-on: ubuntu-latest

    steps:
      - name: Checkout
        uses: actions/checkout@v4

      - name: Setup Rust
        uses: dtolnay/rust-toolchain@stable
        with:
          components: rustfmt

      - name: Cache Rust build artifacts
        uses: Swatinem/rust-cache@v2
        with:
          workspaces: diffly-rust

      - name: Check Rust formatting
        run: cargo fmt --manifest-path diffly-rust/Cargo.toml --all --check

      - name: Run Rust unit tests
        run: cargo test --manifest-path diffly-rust/Cargo.toml

      - name: Run Rust conformance runner
        run: make test-spec-rust

      - name: Run Rust engine conformance runner (multi-partition parity)
        run: make test-spec-rust-engine PARTITIONS=4

      - name: Run Rust CLI smoke test
        run: |
          make diff-rust \
            A=diffly-spec/fixtures/keyed_basic_add_remove_change/a.csv \
            B=diffly-spec/fixtures/keyed_basic_add_remove_change/b.csv \
            KEY=id > /tmp/diffly-rust-smoke.jsonl

      - name: Run Rust positional CLI smoke test (default mode)
        run: |
          make diff-rust \
            A=diffly-spec/fixtures/positional_basic_add_remove_change/a.csv \
            B=diffly-spec/fixtures/positional_basic_add_remove_change/b.csv > /tmp/diffly-rust-smoke-positional.jsonl

      - name: Run Rust unordered CLI smoke test (--ignore-row-order)
        run: |
          make diff-rust \
            A=diffly-spec/fixtures/positional_ignore_row_order_basic_add_remove/a.csv \
            B=diffly-spec/fixtures/positional_ignore_row_order_basic_add_remove/b.csv \
            IGNORE_ROW_ORDER=1 > /tmp/diffly-rust-smoke-unordered.jsonl

      - name: Run Rust ignore-column-order keyed smoke test
        run: |
          make diff-rust \
            A=diffly-spec/fixtures/keyed_header_sorted_mode_add/a.csv \
            B=diffly-spec/fixtures/keyed_header_sorted_mode_add/b.csv \
            KEY=id \
            IGNORE_COLUMN_ORDER=1 > /tmp/diffly-rust-smoke-ignore-column-order.jsonl

      - name: Verify Rust smoke output has schema and stats
        run: |
          python3 - <<'PY'
          import json
          from pathlib import Path

          lines = [
              json.loads(line)
              for line in Path("/tmp/diffly-rust-smoke.jsonl").read_text(encoding="utf-8").splitlines()
              if line.strip()
          ]
          assert lines, "Rust smoke output is empty"
          assert lines[0].get("type") == "schema", "First event must be schema"
          assert lines[-1].get("type") == "stats", "Last event must be stats"

          positional_lines = [
              json.loads(line)
              for line in Path("/tmp/diffly-rust-smoke-positional.jsonl").read_text(encoding="utf-8").splitlines()
              if line.strip()
          ]
          assert positional_lines, "Rust positional smoke output is empty"
          assert positional_lines[0].get("type") == "schema", "First positional event must be schema"
          assert positional_lines[-1].get("type") == "stats", "Last positional event must be stats"

          unordered_lines = [
              json.loads(line)
              for line in Path("/tmp/diffly-rust-smoke-unordered.jsonl").read_text(encoding="utf-8").splitlines()
              if line.strip()
          ]
          assert unordered_lines, "Rust unordered smoke output is empty"
          assert unordered_lines[0].get("type") == "schema", "First unordered event must be schema"
          assert unordered_lines[-1].get("type") == "stats", "Last unordered event must be stats"
          assert unordered_lines[-1].get("rows_changed") == 0, "Rust unordered rows_changed must be 0"

          ignore_column_lines = [
              json.loads(line)
              for line in Path("/tmp/diffly-rust-smoke-ignore-column-order.jsonl").read_text(encoding="utf-8").splitlines()
              if line.strip()
          ]
          assert ignore_column_lines, "Rust ignore-column-order smoke output is empty"
          assert ignore_column_lines[0].get("type") == "schema", "First ignore-column-order event must be schema"
          assert ignore_column_lines[-1].get("type") == "stats", "Last ignore-column-order event must be stats"
          PY

      - name: Run Rust CLI partitioned smoke test
        run: |
          make diff-rust \
            A=diffly-spec/fixtures/keyed_basic_add_remove_change/a.csv \
            B=diffly-spec/fixtures/keyed_basic_add_remove_change/b.csv \
            KEY=id \
            PARTITIONS=4 > /tmp/diffly-rust-smoke-partitioned.jsonl

      - name: Verify partitioned smoke output has schema and stats
        run: |
          python3 - <<'PY'
          import json
          from pathlib import Path

          lines = [
              json.loads(line)
              for line in Path("/tmp/diffly-rust-smoke-partitioned.jsonl").read_text(encoding="utf-8").splitlines()
              if line.strip()
          ]
          assert lines, "Partitioned rust smoke output is empty"
          assert lines[0].get("type") == "schema", "First event must be schema"
          assert lines[-1].get("type") == "stats", "Last event must be stats"
          PY

  web-app:
    name: Web App Checks
    runs-on: ubuntu-latest
    needs:
      - python-reference
      - rust-conformance

    steps:
      - name: Checkout
        uses: actions/checkout@v4

      - name: Setup Node
        uses: actions/setup-node@v4
        with:
          node-version: "22"
          cache: "npm"
          cache-dependency-path: diffly-web/package-lock.json

      - name: Install web dependencies
        run: npm --prefix diffly-web ci

      - name: Lint web app
        run: make web-lint

      - name: Typecheck web app
        run: make web-typecheck

      - name: Build web app
        run: npm --prefix diffly-web run build

  deploy-firebase-hosting:
    name: Deploy Firebase Hosting (prod)
    runs-on: ubuntu-latest
    needs: web-app
    if: github.ref == 'refs/heads/main' && (github.event_name == 'push' || github.event_name == 'workflow_dispatch')
    environment: production

    steps:
      - name: Checkout
        uses: actions/checkout@v4

      - name: Setup Node
        uses: actions/setup-node@v4
        with:
          node-version: "22"
          cache: "npm"
          cache-dependency-path: diffly-web/package-lock.json

      - name: Install web dependencies
        run: npm --prefix diffly-web ci

      - name: Build web app
        run: npm --prefix diffly-web run build

      - name: Install Firebase CLI
        run: npm install -g firebase-tools

      - name: Deploy to Firebase Hosting
        env:
          FIREBASE_PROJECT_ID: ${{ secrets.FIREBASE_PROJECT_ID }}
          FIREBASE_TOKEN: ${{ secrets.FIREBASE_TOKEN }}
        run: |
          if [ -z "$FIREBASE_PROJECT_ID" ]; then
            echo "Missing required secret: FIREBASE_PROJECT_ID"
            exit 1
          fi
          if [ -z "$FIREBASE_TOKEN" ]; then
            echo "Missing required secret: FIREBASE_TOKEN"
            exit 1
          fi
          firebase deploy --only hosting --project "$FIREBASE_PROJECT_ID" --token "$FIREBASE_TOKEN"